What's new here is the mechanism: interviews with 15 drivers show algorithms hiring, rating, and firing contract drivers with little or no human review — extending the pattern drivers already reported with the Mentor driving-score app, whose bugs they said triggered unfair disciplinary action.
First-order effects
Flex contractors face termination, deactivation, and hiring decisions made by automated systems with no human appeal path, meaning a bad algorithmic rating can end their income directly.
Amazon gains labor flexibility at near-zero management cost — no human managers are in the loop for the hire-rate-fire cycle across its contract driver pool.
Second-order effects
Driver grievances shift from pay disputes (the 2018 wage shortfall, the 2019 tips controversy) to contestability of automated decisions, giving organizers and investigators a new concrete target: opaque scoring systems like Mentor.
Regulators and plaintiffs' attorneys get documented cases of fully automated adverse employment actions against gig workers, pressuring platforms to add human review or disclosure.
Third-order effects
If the pattern holds, algorithmic management becomes the default employment relationship for gig work — hiring, evaluation, and firing handled end-to-end by software, with workers classified as contractors precisely so no employer-side HR function has to exist.
The trend: Gig platforms are replacing human managers with end-to-end algorithmic management, moving disputes over wages and fairness into disputes over automated decision-making.
You just wait until the headquarters AI is writing your evaluation based on an HRC productivity algorithm. No amount of sucking up to the T-800 in the command suite is going to save you then. https://www.bloomberg.com/...
@jensmithwsj
Jennifer Smith on x
Amazon knew delegating work to machines would lead to mistakes and damaging headlines, former managers said, but decided it was cheaper to trust the algorithms than pay people to investigate mistaken firings so long as the drivers could be replaced easily. https://www.bloomberg.c…
@spencersoper
Spencer Soper on x
Worried a robot will take your job? What if a robot became your boss and fired you? Welcome to the present at #amazon, where millions of delivery drivers are managed by machines. https://www.bloomberg.com/...
@johnkoetsier
John Koetsier on x
Amazon is arbitrarily firing workers by algorithm for reasons that make zero human sense. :-( You are harming human beings with lives, dependents, obligations, feelings, and emotions, @JeffBezos. This is evil. Be better. https://www.bloomberg.com/...
@ethanz
Ethan Zuckerman on x
Scholars think a great deal about algorithmic fairness. But what about when the authors of algorithms don't give a crap about fairness, concluding that screwing over edge cases is more profitable than hiring more management? Fascinating article re: Amazon: https://www.bloomberg.c…
@kyweise
Karen Weise on x
“Executives knew this was gonna shit the bed,” this person said. “That's actually how they put it in meetings. The only question was how much poo we wanted there to be.” 👀 @spencersoper's new story on the automated firing of Amazon's Flex drivers https://www.bloomberg.com/...
@mattdpearce
Matt Pearce on x
Job-stealing robots have come for managers' jobs! https://www.bloomberg.com/...
@mhbergen
Mark Bergen on x
“Normandin's experience is a twist on the decades-old prediction that robots will replace workers.” From the inimitable @spencersoper https://www.bloomberg.com/...
@johnjedwards3
John J. Edwards III on x
Fascinating investigative story here. Amazon uses algorithms to rate and sometimes fire contract workers, knowing the bots will make mistakes but figuring the extra overall efficiency is worth it https://www.bloomberg.com/... via @technology @spencersoper
@hypervisible
Hypervisible Dot Pdf on x
“At Amazon, machines are often the boss—hiring, rating and firing millions of people with little or no human oversight.” https://www.bloomberg.com/...